
/* use the file, "da100.dta" to produce the models from columns 1 and 2 of Table 2 and Figure 1 in the article */

use "filepath\da100.dta"

/* Table 2, column 1; fixed effects estimated directly, obviously */

replace pages=pages/100

logit d_oversight1 nagency nwitnesses days pages dum_sc1- dum_sc5

/* Figure 1 */

margins, at(nagency=(0 1 2 3 4 5 6 7 8 9 10 11 12 13))

#delimit ;

marginsplot,plot1opts(lpattern(solid) color(black)  )
ylabel(0(.25)1) xlabel(0(2)12)
xtitle("No. of Exec. Branch Personnel Testifying",height(5) size(small)) ytitle("Prob.  Primary Purpose Is Oversight",height(5))
title("Figure 1. The Estimated Influence of the Number of Executive Personnel on the Probability",color(black) size(small) )
subtitle("that a Hearing's Primary Purpose Is Oversight, with 95% Confidence Intervals",color(black) size(small) )
saving("[Insert Filepath\filename]")  ;

#delimit cr

/* Table 2, column 2; fixed effects estimated directly */

logit d_oversight2 nagency nwitnesses days pages dum_sc1- dum_sc5

clear

/* use the file, "da91_115" to produce the models from columns 3 and 4 of Table 2 in the article */

use "filepath\da91_115"

replace pages=pages/100

/* Note I do not include an observation with 52 executive branch personnel testifying. Includling
it does not change the findings */

/* Column 3, Table 2,  fixed effects estimated directly */

logit d_oversight1 nagency nwitnesses days pages c91-c102 c104-c115 p1-p6 if nagency~=52

/* Column 4, Table 2,  fixed effects estimated directly */

logit d_oversight2 nagency nwitnesses days pages c91-c102 c104-c115 p1-p6 if nagency~=52


